Expandable TLIF cage subsidence impact on radiographic and clinical outcomes

2017

Summary

Clinical and radiologic outcomes comparing TLIF cage subsidence vs no subsidence in a prospective outcomes study of 177 consecutive adults undergoing posterior fusion with TLIF using an expandable cage. Subsidence was more common in women, lumbar hypolordosis, and at L4-5. Subsidence was associated with average 2mm less anterior disc height, 3? loss of disc lordosis, and 4? loss of lumbar lordosis at 2 years. Subsidence did not appear to affect 2-year clinical outcomes or complications.

Hypothesis

TLIF cage subsidence has negative clinical and radiographic impact

Design

Prospective outcomes study from consecutive series at one center

Introduction

Small series comparing expandable TLIF cage subsidence vs no subsidence suggested no clinical difference, though a larger series was needed to clarify these findings. Radiographic outcomes with subsidence have not been reported.

Methods

Prospective data was collected on adults with posterior fusion and TLIF using expandable cage from 2010-2015, minimum 24mo follow-up. Patients were divided into two groups: cage subsidence vs no subsidence. Subsidence defined: >2mm cage settling on standing lateral radiographs. Clinical and radiographic data were collected preop, at 1year and 2 years.

Results

177 patients age 65 yrs (19-84yrs); smokers-10%, BMI-29.7(19-48), 41% had prior surgery. Posterior fusion averaged 4.0 levels (1–16 levels); TLIF at 304 levels (1-level: 68, 2-level:90, 3-level: 15, 4-level: 3); Subsidence occurred in 40 (22.5%) patients, more in females(p=0.08). Deformity patients (63) had less subsidence than degenerative (62.5%vs37.5%; p<0.001), but 40 had iliac screws. Subsidence was most common at L4-L5(57.5%; p< 0.001), at inferior endplate(p<0.001), with average 2mm loss of anterior disc height, 3? loss of disc lordosis, and 4? loss of lumbar lordosis. At 2 years, T12-S1 improved 8? without subsidence, 3? with subsidence (p<0.05). Subsidence was more common with less lumbar lordosis (p = 0.004) but was unaffected by pelvic incidence. At 2 years, pain medication scores, VAS-back, VAS-leg, ODI, and ODI walking-scores improved with surgery (p<0.01) but were unaffected by subsidence. Complications including neuro deficit (5), nonunion (2), hardware failure (1), and need for revision surgery(19) were also unaffected.

Conclusions

Expandable cage subsidence in TLIF was more common at the inferior endplate, at L4-5, in women, in hypolordotic lumbar spines, and in fusions for degenerative disease without spinopelvic fixation. Subsidence was associated with average 2mm less anterior disc height, 3? loss of disc lordosis, and 4? loss of lumbar lordosis at 2 years. Subsidence did not appear to affect 2-year clinical outcomes or complications.